You are on page 1of 6

laboratory science

Blue light-absorbing intraocular lens and retinal

pigment epithelium protection in vitro
Janet R. Sparrow, PhD, Ashley S. Miller, Jilin Zhou, MD

Purpose: To compare the Alcon AcrySof威 Natural (SN60AT) and AcrySof

(SA60AT), the AMO Sensar威 (AR40e) and ClariFlex威, and the Pfizer CeeOn威 Edge
911A intraocular lenses (IOLs) as to their ability to protect retinal pigment epithe-
lial (RPE) cells from light damage mediated by the lipofuscin fluorophore A2E.
Setting: Department of Ophthalmology, Columbia University, New York, New
York, USA.
Methods: Cultured human RPE cells (ARPE-19 cell line) that had accumulated
A2E were exposed to blue (430 nm ⫾ 30), green (550 ⫾ 10 nm), or white (390 to
750 nm) light with and without an IOL in the light path.
Results: The blue light-absorbing AcrySof Natural IOL was associated with sig-
nificant reduction (78% to 82%; P⬍.01) in the death of A2E-laden RPE that were
exposed to blue, white, and green light. The decrease in the incidence of cell
death was greater in magnitude than would be expected from the amount of light
that was absorbed by the IOL. The considerably smaller declines in cell death ob-
served with the AcrySof, Sensar, ClariFlex, and CeeOn Edge IOLs were likely due
to nonspecific reductions in light transmittance.
Conclusions: By absorbing blue light, the AcrySof Natural IOL shields RPE cells
that have accumulated the aging lipofuscin fluorophore A2E from the damaging ef-
fects of light. A long-term population-based clinical trial would determine whether
a blue light-absorbing IOL can reduce the risk for or progression of age-related
macular degeneration.
J Cataract Refract Surg 2004; 30:873–878  2004 ASCRS and ESCRS

T he design of intraocular lenses (IOLs) should be

based on the properties of the human ocular lens,
especially in the transmission properties of the IOL.
human lens, since the latter yellows with age while IOLs
in current use are colorless.2,9–13 The color change in the
natural lens is likely attributable to oxidation products of
Notably, the human crystalline lens absorbs most ultra- tryptophan (n-formyl-kynurenine) and to glycosylation
violet light between 300 nm and 400 nm.1 Thus, to of lens proteins.14 It results in a progressive increase
provide the same protection to the retina afforded by in absorbance within the blue range of the visible
the natural lens, the use of ultraviolet light-absorbing spectrum.2,9
IOLs became the standard of practice once these lenses Filtering the shorter wavelengths of the visible spec-
were developed.2–8 However, the transmission properties trum is particularly significant because it is also this
of most IOLs are not comparable to those of the natural portion of the spectrum that produces photochemical
damage to the retinal pigment epithelium (RPE).15–17
It is now generally accepted that at least 1 of the intracel-
Accepted for publication January 27, 2004.
lular chromophores responsible for the blue light sensi-
Reprint requests to Janet R. Sparrow, PhD, Department of Ophthalmol-
ogy, Columbia University, 630 West 128th Street, New York, New tivity of RPE cells18,19 is the lipofuscin constituent
York 10032, USA. email: A2E.20–22 This fluorophore is unique to RPE cells and
 2004 ASCRS and ESCRS 0886-3350/04/$–see front matter
Published by Elsevier Inc. doi:10.1016/j.jcrs.2004.01.031

has been shown to accumulate in human RPE cells CeeOn威 Edge (911A) (silicone, 20.0 D, 6.0 mm optic
throughout the lifetime of an individual, with levels diameter).
being highest in the aged eye.21 Thus, replacement of
a senile cataractous crystalline lens with a colorless IOL A2E Accumulation in Culture
may leave the RPE vulnerable at an age when its content Human RPE cells (ARPE-19, American Type Culture
of blue light sensitive A2E is already high. Collection), which are devoid of endogenous A2E,35 were
grown in 8-well, plastic chamber slides (Laboratory-Tek,
The RPE fluorophore A2E is maximally excited
Nunc), as described.35 Once confluent, the cells were allowed
by light in the blue region of the spectrum.18 When to accumulate synthesized A2E21 from a 20 ␮M concentration
irradiated, A2E generates singlet oxygen that proceeds to added to the medium. With this protocol, A2E accumulates
add to carbon–carbon double bonds of A2E to generate in the lysosomal compartment of the cells to levels that are
highly reactive epoxides (A2E-epoxides) along the side- comparable to amounts present in vivo.18,35
arms of the molecule.23, 24 The cellular injury induced by
the illumination of A2E-laden RPE includes oxidative Illumination and Placement of the IOL
DNA base changes,25,26 and it is likely that as electro- Immediately before illumination, the culture medium
philes that can readily react with many cellular mole- was replaced with phosphate-buffered saline containing cal-
cules, A2E-epoxides may account for much of the cium, magnesium, and glucose. For quantitative measure-
cellular damage accrued.25 The photochemical events ments of light-induced cell death, the cells were exposed to
provoked by the irradiation of A2E-laden RPE ulti- blue light (430 nm ⫾ 30 [SD], 8 mW/cm2), green light
(550 ⫾ 10 nm, 8 mW/cm2), or white light (246 mW/cm2)
mately result in the initiation of a cell death program.
over a 0.8 mm ⫻ 8.5 mm field. The light was delivered
The sensitivity to blue light conferred by the lipofuscin
from a tungsten halogen source for 20 minutes, and power
fluorophore A2E may explain why atrophic age-related was measured with a Newport optical power meter (model
macular degeneration (AMD) has been linked to both 840). The spectral range of the lamp was 390 to 750 nm,
RPE lipofuscin27–33 and cumulative light exposure.34 with a lower output at the shorter wavelengths.
Moreover, the loss of RPE cells in atrophic AMD is a The IOL to be tested was applied to the undersurface
critical event as it leads to photoreceptor cell degeneration. of the culture well, where it remained attached, unaided. It
Given that the attenuation of blue light afforded was centered over the light path.
by the yellowed senescent lens may defend lipofuscin- To obtain a visual representation of IOL protection,
filled RPE cells against blue light damage, we are in- 1.0 mm diameter disks were cut from the center of the IOL
using a trephine blade (Katena Products, Inc.). The IOL
terested in efforts being made to develop IOLs that
disks were positioned on the undersurface of the well, which
replicate the transmission characteristics of the aging
was then irradiated (430 ⫾ 30 nm, 16 mW/cm2) from below.
crystalline lens. In this study, we constructed a cell
culture system that allowed us to compare several IOLs
Detection of Nonviable Cells
as to their ability to protect A2E-laden RPE from blue
light damage. In this cell culture system, A2E accumu- The nuclei of dead RPE cells were labeled with a mem-
brane impermeant dye (Dead Red, Molecular Probes), and
lates in the lysosomal compartment of a human RPE the nuclei of all cells with 4⬘,6⬘-diamino-2-phenylindole
cell line, as it does in RPE of the eye.35 Moreover, the (DAPI), as reported.25,26 Blue light-illuminated A2E-con-
levels of A2E are comparable to the level occurring in taining RPE labeled in this way are undergoing an apoptotic
vivo.18,35 This model also allows us to study RPE cells form of cell death.36 Digital images (5 fields per illumination
with and without intracellular deposits of A2E. zone) were obtained using a Zeiss Axioplan II microscope
with Axiocam camera and KS400 image processing software.
Subsequently, Dead Red and DAPI-stained nuclei were
Materials and Methods counted. Dead cells were quantified as a percentage of the
The IOLs studied were the Alcon AcrySof威 Natural total number of cells in a field. Means are based on 3
(SN60AT) (acrylic, 20.0 diopters [D], 6.0 mm optic diame- experiments.
ter) and AcrySof (SA60AT) (acrylic, 20.0 D, 6.0 mm optic Statistical analysis was by an analysis of variance followed
diameter); the AMO ClariFlex威 (CLRFLXB) (silicone, by the Newman-Keul multiple comparison test (Prism,
20.0 D, 6.0 mm optic diameter) and Sensar威 (AR40e) GraphPad Software). A P value of 0.05 or less was consid-
(acrylic, 20.0 D, 6.0 mm optic diameter); and the Pharmacia ered significant.



that did not contain A2E (no A2E, no IOL) remained

viable, while blue light-illuminated (430 nm peak with
a bandwidth of 60 nm, 8 mW/cm2) RPE that had
previously accumulated A2E underwent marked cell
death. In the present experiments, 41.1% ⫾ 4.1% (3
experiments) of the A2E-laden cells in a field of illumi-
nation became nonviable after blue light exposure in
the absence of an IOL (Figure 1). When the yellow-
tinted AcrySof Natural IOL was placed in the center
of the light path, transmission of the 430 nm light
was reduced by approximately 50% and cell death was
reduced by 80% (P⬍.001) compared to irradiation in
the absence of an IOL (Figure 1). When the AcrySof,
Sensar, ClariFlex, or CeeOn Edge IOLs were in place,
Figure 1. (Sparrow) Quantitation of nonviable RPE cells after A2E the frequency of nonviable cells was not different
accumulation and blue light illumination (430 nm) with and without (P⬎.05) than cell death in the absence of an IOL. The
an IOL placed in the light path. The percentage of nonviable cells
modest but consistent declines in cell death that were
was determined by labeling all nuclei with DAPI and the nuclei of
nonviable cells with a membrane-impermeable dye. The data were detected with these IOLs (Figure 1) were undoubtedly
normalized to the percentage of nonviable cells in the absence of an due to the small reductions in light transmission (⬃5%)
IOL. Values are mean ⫾ SEM. that were measured.
The ability of the AcrySof Natural IOL to attenuate
blue light-mediated cell death was evident when fields
of irradiated cells were imaged. In the absence of an IOL,
the nuclei of nonviable cells were uniformly distributed
across the illuminated field (Figure 2). Conversely,
placement of the blue light-absorbing AcrySof Natural
IOL (a 1.0 mm disk) within the light path spared a
circular zone of cells, the diameter of this area correspond-
ing to the diameter of the IOL disc. Within this sector,
all nuclei were labeled with DAPI but labeled nuclei
of nonviable cells were only occasionally observed.
With the use of wide-band white light (390 to
750 nm; 246 mW/cm2) that included all wavelengths
toward which A2E exhibits sensitivity, the death of
A2E-laden RPE reached a frequency of 35% (35.4% ⫾
Figure 2. (Sparrow) A blue light-absorbing IOL protects A2E- 2.7% of the cells in an illumination zone; 3 experi-
laden RPE from cell death due to 430 nm irradiation. The nuclei of ments). In the presence of the blue light-absorbing Acry-
nonviable RPE were labeled with a membrane impermeable dye (left), Sof Natural IOL, cell death from white light was
and all nuclei were stained with DAPI (right). The left and right panels
reduced by 82% (P⬍.001) compared to that without
are corresponding fields of illuminated A2E-laden RPE. In the absence
of an IOL in the light path (top), the entire field is covered with nonviable an IOL (Figure 3). In comparison, the AcrySof IOL
cells. Placement of a blue light-absorbing IOL (1.0 mm diameter disk; diminished the incidence of nonviable cells by 22%
AcrySof Natural) in the light path reduced the number of nonviable (P⬍.05); with the Sensar, ClariFlex, and CeeOn Edge
cells in a circular zone corresponding to the area covered by the IOL
IOLs, cell death was decreased by 37% to 39% (P⬍
(bottom, dotted circle).
.01). Again, the decreases observed with the colorless
IOLs were likely related to the reductions in spectral
Results transmittance associated with all IOLs.
Using a fluorescence assay that labels the nuclei of When compared with blue and white light, green
nonviable cells, it was observed that illuminated RPE light (550 ⫾ 20 nm) delivered at an intensity that was



Figure 3. (Sparrow) Quantitation of nonviable cells after white light Figure 4. (Sparrow) Quantitation of nonviable cells after green light
(390 to 750 nm) illumination of A2E-laden RPE with and without the illumination (550 nm) of A2E-laden RPE with and without the indicated
indicated IOL placed in the light path. The percentage of nonviable IOL placed in the light path. Values for percentage of nonviable cells
cells was normalized to values obtained in the absence of an IOL. were normalized to values obtained in the absence of an IOL. Means ⫾
Values are mean ⫾ SEM. SEM are presented.

the same as that of the blue light (8 mW/cm2) induced with the other IOLs were probably due to IOL-associ-
the death of only a small number of A2E-laden RPE ated nonspecific decreases (5% to 10%) in light trans-
cells (2.9% ⫾ 0.6%; 3 experiments). With green light mission that have been reported.2,7,8,37
irradiation, all IOLs reduced the death of RPE that had The AcrySof Natural IOL did not completely abol-
accumulated A2E (Figure 4). With the AcrySof Natural, ish cell death in this model. However, this is not surpris-
the number of nonviable cells was decreased by 78% ing since by design, this IOL does not block all blue
(P⬍.01) compared to cell death without an IOL; while light. Rather, the yellow tint of the lens is intended to
cell death declined by 33% with the AcrySof (P⬎.05) confer a spectral transmission similar to that of the
and by 52% to 57% (P⬍.05) with the Sensar, ClariFlex, crystalline lens of an adult.1 That A2E-containing RPE
and CeeOn Edge IOLs. are less sensitive than to green light is to be expected:
A2E has an excitation maximum of approximately
430 nm with the emission elicited at 550 nm being less
Discussion than 5% of that elicited at 430 nm.18
The study shows that by absorbing predominantly Ham et al.15 were the first investigators to demon-
in the blue region of the spectrum, the AcrySof Natural strate that retinal injury from wavelengths that peak
IOL shields RPE cells that have accumulated the lipofus- around 425 nm is initiated by photochemical processes
cin fluorophore A2E from the damaging effects of light. in the RPE. The action spectrum of this damage corre-
The protection afforded by the AcrySof Natural IOL sponds to the absorption spectrum of the aging fluo-
was pronounced for wide-band white light just as it rophore A2E,18 and there is now an abundance of
was for narrow-band blue light. The reduction in cell experimental evidence that A2E is a sensitizing molecule
death afforded by this blue light-filtering IOL was that can mediate light injury to RPE, leading to RPE
greater in magnitude than would be expected from the atrophy.18,23,25,26,36 Since lipofuscin fluorophores, includ-
decline in transmitted blue light. That a reduction in ing A2E, accumulate with age, the natural yellowing
blue light transmission of approximately 50% resulted of the aging human lens is fortunate as it may dampen
in an 80% decline in cell death probably occurred be- the damaging potential of short wavelength blue light.
cause blue light levels were brought below the threshold For some time, it has been postulated that cumula-
for lethal damage for a significant proportion of the tive photochemical damage from lifetime exposures to
cells. Small but consistent declines in cell death observed light are a cause of AMD,38 although the results of



epidemiological studies are inconclusive.39,40 Neverthe- 5. Lawrence HM, Reynolds TR. Erythropsial phototoxi-
less, it is of interest that the Chesapeake Bay Waterman city associated with nonultraviolet-filtering intraocular
lenses. J Cataract Refract Surg 1989; 15:569–572
Study reported an association between the incidence of
6. Peyman GA, Zak R, Sloane H. Ultraviolet-absorbing
advanced AMD and blue light exposure during the 20- pseudophakos: an efficacy study. Am Intra-Ocular Im-
year period that preceded the study.34 Other reports plant Soc J 1983; 9:161–170
suggest a relationship between the incidence of AMD 7. Hammer HM, Yap M, Weatherill JR. Visual perfor-
and cataract extraction. For instance, the National mance in pseudophakia with standard and ultraviolet-
Health and Nutrition Survey observed 3087 patients absorbing intraocular lenses: a preliminary report. Trans
between 1970 and 1974 and reports that aphakic sub- Ophthalmol Soc U K 1986; 105:441–446
8. Mainster MA. The spectra, classification, and rationale
jects had an increased risk for developing AMD.41 In
of ultraviolet-protective intraocular lenses. Am J Oph-
a study of 47 patients who presented with bilateral thalmol 1986; 102:727–732
drusen and pigmentary changes, choroidal neovasculari- 9. Mellerio J. Yellowing of the human lens: nuclear and
zation was observed in 19.1% of the eyes that had had cortical contributions. Vis Res 1987; 27:1581–1587
cataract extraction with implantation of a colorless IOL 10. Weale RA. Human lenticular fluorescence and transmis-
compared with an incidence of 4.3% in the fellow sivity, and their effects on vision. Exp Eye Res 1985;
phakic eyes.42 The Beaver Dam Study comprised 4926 41:457–473
11. Gaillard ER, Zheng L, Merriam JC, Dillon J. Age-related
patients from 1988 to 1990 and found a positive associa-
changes in the absorption characteristics of the primate
tion between cataract extraction and clinical indications lens. Invest Ophthalmol Vis Sci 2000; 41:1454–1459
of early AMD.43 Indeed, both the 5-year and 10-year 12. Weale RA. Aging and vision. Vis Res 1986; 26:1507–
follow-up studies confirm that eyes that had had cataract 1512
surgery before baseline evaluation were more likely to 13. Weale RA. Age and the transmittance of the human
exhibit progression of AMD.44,45 Nevertheless, similar crystalline lens. J Physiol 1988; 395:577–587
associations were not reported by the Age-Related Eye 14. Chylack LT Jr. Aging changes in the crystalline lens and
zonules. In: Albert DM, Jakobiec FA, eds, Principles and
Disease Study (DF Martin, ARVO abstract 1907,
Practice of Ophthalmology. Basic Sciences. Philadelphia,
2002). PA, WB Saunders Co, 1994; 702–710
Circumstances that lead to damage in the aging RPE 15. Ham WT Jr, Mueller HA, Ruffolo JJ Jr, et al. Basic
cell are considered as a prelude to the photoreceptor mechanisms underlying the production of photochemi-
cell degeneration that characterizes the visual impairment cal lesions in the mammalian retina. Curr Eye Res
associated with AMD. We speculate that a yellow-tinted 1984; 3:165–174
IOL that simulates the adult natural lens and protects 16. Ham WT Jr, Ruffolo JJ Jr, Mueller HA, et al. Histologic
lipofuscin-containing RPE from blue light damage may analysis of photochemical lesions produced in rhesus
retina by short-wavelength light. Invest Ophthalmol Vis
reduce the risk for or progression of AMD. This concept
Sci 1978; 17:1029–1035
warrants evaluation by a long-term population-based 17. Ham WT Jr, Allen RG, Feeney-Burns L, et al. The
clinical study. involvement of the retinal pigment epithelium (RPE).
In: Waxler M, Hitchens VM, eds, Optical radiation and
References Visual Health. Boca Raton, FL, CRC Press, 1986; 43–67
1. Boettner EA, Wolter JR. Transmission of the ocular 18. Sparrow JR, Nakanishi K, Parish CA. The lipofuscin
media. Invest Ophthalmol 1962; 1:776–783 fluorophore A2E mediates blue light-induced damage to
2. Lindstrom RL, Doddi N. Ultraviolet light absorption retinal pigmented epithelial cells. Invest Ophthalmol Vis
in intraocular lenses. J Cataract Refract Surg 1986; Sci 2000; 41:1981–1989
12:285–289 19. Schütt F, Davies S, Kopitz J, et al. Photodamage to
3. Komatsu M, Kanagami S, Shimizu K. Ultraviolet- human RPE cells by A2-E, a retinoid component of
absorbing intraocular lens versus non-UV-absorbing lipofuscin. Invest Ophthalmol Vis Sci 2000; 41:2303–
intraocular lens: comparison of angiographic cystoid 2308
macular edema. J Cataract Refract Surg 1989; 15:654– 20. Eldred GE, Lasky MR. Retinal age pigments generated
657 by self-assembling lysosomotropic detergents. Nature
4. Werner JS, Hardenbergh FE. Spectral sensitivity of the 1993; 361:724–726
pseudophakic eye. Arch Ophthalmol 1983; 101:758– 21. Parish CA, Hashimoto M, Nakanishi K, et al. Isolation
760 and one-step preparation of A2E and iso-A2E, fluoro-



phores from human retinal pigment epithelium. Proc 34. Taylor HR, West S, Muñoz B, et al. The long-term
Natl Acad Sci USA 1998; 95:14609–14613 effects of visible light on the eye. Arch Ophthalmol
22. Sakai N, Decatur J, Nakanishi K, Eldred GE. Ocular 1992; 110:99–104
age pigment “A2E”: an unprecedented pyridinium bis- 35. Sparrow JR, Parish CA, Hashimoto M, Nakanishi K.
retinoid. J Am Chem Soc 1996; 118:1559–1560 A2E, a lipofuscin fluorophore, in human retinal pig-
23. Sparrow JR, Zhou J, Ben-Shabat S, et al. Involvement mented epithelial cells in culture. Invest Ophthalmol Vis
of oxidative mechanisms in blue-light-induced damage Sci 1999; 40:2988–2995
to A2E-laden RPE. Invest Ophthalmol Vis Sci 2002; 36. Sparrow JR, Cai B. Blue light-induced apoptosis of A2E-
43:1222–1227 containing RPE: involvement of caspase-3 and protec-
24. Ben-Shabat S, Itagaki Y, Jockusch S, et al. Formation tion by Bcl-2. Invest Ophthalmol Vis Sci 2001; 42:
of a nonaoxirane from A2E, a lipofuscin fluorophore 1356–1362
related to macular degeneration, and evidence of singlet 37. Mainster MA. Spectral transmittance of intraocular
oxygen involvement. Angew Chem Int Ed Engl 2002; lenses and retinal damage from intense light sources. Am
41:814–817 J Ophthalmol 1978; 85:167–170
25. Sparrow JR, Vollmer-Snarr HR, Zhou J, et al. A2E- 38. Mainster MA. Light and macular degeneration: a bio-
epoxides damage DNA in retinal pigment epithelial cells. physical and clinical perspective. Eye 1987; 1:304–310
Vitamin E and other antioxidants inhibit A2E-epoxide 39. Darzins P, Mitchell P, Heller RF. Sun exposure and age-
formation. J Biol Chem 2003; 278:18207–18213 related macular degeneration; an Australian case-control
26. Sparrow JR, Zhou J, Cai B. DNA is a target of the study. Ophthalmology 1997; 104:770–776
photodynamic effects elicited in A2E-laden RPE by blue- 40. Cruickshanks KJ, Klein R, Klein BEK. Sunlight and
light illumination. Invest Ophthalmol Vis Sci 2003; 44: age-related macular degeneration; the Beaver Dam
2245–2251 Study. Arch Ophthalmol 1993; 111:514–518
27. Weiter JJ, Delori FC, Wing GL, Fitch KA. Retinal 41. Liu IY, White L, LaCroix AZ. The association of age-
pigment epithelial lipofuscin and melanin and choroidal related macular degeneration and lens opacities in the
melanin in human eyes. Invest Ophthalmol Vis Sci aged. Am J Public Health 1989; 79:765–769
1986; 27:145–152 42. Pollack A, Marcovich A, Bukelman A, Oliver M. Age-
28. Wing GL, Blanchard GC, Weiter JJ. The topography related macular degeneration after extracapsular cataract
and age relationship of lipofuscin concentration in the extraction with intraocular lens implantation. Ophthal-
retinal pigment epithelium. Invest Ophthalmol Vis Sci mology 1996; 103:1546–1554
1978; 17:601–607 43. Klein R, Klein BEK, Wang Q, Moss SE. Is age-related
29. Delori FC, Dorey CK, Staurenghi G, et al. In vivo maculopathy associated with cataracts? Arch Ophthal-
fluorescence of the ocular fundus exhibits retinal pigment mol 1994; 112:191–196
epithelium lipofuscin characteristics. Invest Ophthalmol 44. Klein R, Klein BEK, Jensen SC, Cruickshanks KJ. The
Vis Sci 1995; 36:718–729 relationship of ocular factors to the incidence and pro-
30. Delori FC, Goger DG, Dorey CK. Age-related accumu- gression of age-related maculopathy. Arch Ophthalmol
lation and spatial distribution of lipofuscin in RPE of 1998; 116:506–513
normal subjects. Invest Ophthalmol Vis Sci 2001; 42: 45. Klein R, Klein BEK, Wong TY, et al. The association
1855–1866 of cataract and cataract surgery with the long-term inci-
31. von Rückmann A, Fitzke FW, Bird AC. Fundus autoflu- dence of age-related maculopathy. Arch Ophthalmol
orescence in age-related macular disease imaged with a 2002; 120:1551–1558
laser scanning ophthalmoscope. Invest Ophthalmol Vis
Sci 1997; 38:478–486 From the Department of Ophthalmology, Columbia University, New
32. Holz FG, Bellmann C, Margaritidis M, et al. Patterns York, New York 10032, USA.
of increased in vivo fundus autofluorescence in the junc-
None of the authors has a financial or proprietary interest in any
tional zone of geographic atrophy of the retinal pigment product mentioned.
epithelium associated with age-related macular degenera-
tion. Graefes Arch Clin Exp Ophthalmol 1999; 237: Presented at the XXI Congress of the European Society of Cataract and
145–152 Refractive Surgeons, Munich, Germany, September 2003.
33. Holz FG, Bellman C, Staudt S, et al. Fundus autofluores- Supported by grants from Alcon Laboratories, Ft. Worth, Texas; the
cence and development of geographic atrophy in age- National Eye Institute (EY 12951), Bethesda, Maryland; and by un-
related macular degeneration. Invest Ophthalmol Vis Sci restricted funds from Research to Prevent Blindness, New York, New
2001; 42:1051–1056 York, USA.